Skip to main content

What Hana Is (And How To Use Her)

If your team is unsure how to use Hana, start here.

Get Help Early

You do not need to debug onboarding alone. If results are inconsistent, tools are not behaving as expected, or setup feels unclear, reach out early.

Start with Support Channels, FAQs, and Troubleshooting.

Hana in One Sentence​

Hana is a chat-native information agent for Google Chat: she understands team context, retrieves the right information, and helps you execute actions.

If You Remember Only One Thing​

Use Hana when your team needs to turn scattered information into a concrete next step.

Ask Hana to:

  • collect what happened (from thread/space/docs/memory),
  • extract what matters (decisions, risks, blockers),
  • drive what happens next (tasks, reminders, schedules, updates).

To use Hana reliably, users must complete Google OAuth consent when prompted.

  • Approve the required scopes in the consent screen.
  • If Hana says permissions are missing, reauthorize from the permission card in chat.
  • For org rollouts, ensure users complete consent at least once before relying on tool workflows.

What Hana Is Not​

  • Not just a generic Q&A bot.
  • Not a replacement for your systems of record.
  • Not reliable when prompts are vague and history/scope context is missing.

When Hana Works Best​

Use Hana when you need one of these outcomes:

  • Understand: summarize decisions, extract risks, compare options.
  • Retrieve: pull the right context from chat history, docs, memory, and connected sources.
  • Execute: create reminders, tasks, polls, report updates, and scheduling actions.

Use Cases: Personal vs Business​

Hana supports two primary use-case profiles:

  • BUSINESS_ASSISTANT (default): best for team execution, operations, project coordination, docs, tasks, reporting.
  • PERSONAL_COMPANION: best for one-to-one personal assistant style interactions.

Example invocations:

@Hana set hana use case to BUSINESS_ASSISTANT
@Hana set hana use case to PERSONAL_COMPANION

How Hana Fits Across Industries​

Hana’s information-agent model works across verticals because the workflow is the same: gather context, reason over it, and execute next actions.

  • Engineering / Product: summarize standups, extract blockers, track sprint risks, convert discussions into tasks.
  • Sales / RevOps: summarize account threads, track follow-ups, draft meeting prep, maintain pricing and objection-handling memory.
  • Customer Support: consolidate escalations, identify recurring issues, draft response actions, keep resolution playbooks in memory.
  • Marketing / Growth: summarize campaign performance updates, extract decisions from review calls, coordinate launch checklists.
  • Operations / HR / Admin: track policy updates, manage reminders, generate status digests, keep SOP memory current.
  • Leadership / PMO: generate weekly decision summaries, risk registers, cross-team action dashboards, and ownership tracking.

Simple Prompt Formula​

Use this structure:

@Hana + objective + scope + time window + output format

Example:

@Hana summarize launch decisions in this thread from today 9am-5pm as a table with owner, decision, next step.

Invocation Workflow (Visual)​

10 High-Value Prompts​

@Hana summarize this thread and list action items with owners and due dates.
@Hana lookback list blockers mentioned in this space since Monday.
@Hana create tasks from this thread and assign @A and @B with realistic due dates.
@Hana show active sprint summary with blockers grouped by owner.
@Hana remind me tomorrow 9am to send the customer update.
@Hana create a poll: "Ship this Friday?" options: Yes, No, Need one more day.
@Hana compare the two docs above and highlight missing sections.
@Hana summarize my important unread emails from the last 3 days.
@Hana analyze this meeting transcript and produce decisions + open questions.
@Hana generate an image of a clean product hero banner with white background.

Common Mistakes​

  • Prompt is too vague for low-context chats: "what do you think?" (works better when thread history is rich).
  • No scope: not specifying this thread, this space, my emails, or attached doc.
  • No time window for time-sensitive requests.
  • Mixing multiple tool-heavy actions in one message (for example, "create a reminder and schedule a meeting").

Good Practices​

  • Use one thread per topic to keep context clean.
  • Ask for one outcome per message, then iterate in follow-ups.
  • Split action-heavy requests into separate invocations for better stability (for example, send reminder request first, then meeting request).
  • You can send the second invocation immediately in the same thread/space; you do not need to wait for the first response to finish.
  • Include scope + time window + output format in prompts.
  • Use @mentions for people-specific tasks/reminders.
  • Ground Hana early with one or more memory sources (SOPs, policies, project docs).
  • For high-stakes decisions, verify critical facts before execution.

Example:

@Hana summarize decisions in this thread from today 9am-5pm, then list action items as a table with owner and due date.

First Week Rollout Plan​

  1. Pick one team space and use Hana daily for summaries + action tracking.
  2. Add one memory source (SOP, policy, project doc).
  3. Standardize 5 reusable prompts in your team playbook.
  4. Use thread-first communication so context remains clean.

Where To Go Next​

Last updated on by nav